JP2009009884A - Mems switch, and manufacturing method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a manufacturing method of an MEMS switch capable of improving a switching characteristic and reliability; and an MEMS switch. <P>SOLUTION: A sacrifice layer arranged between a movable electrode 5 and an attraction electrode 4 constituting this MEMS switch is formed by being divided into a first sacrifice layer 8 and a second sacrifice layer 9 formed of materials different from each other, and a region where both of them are superposed on each other is formed. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a manufacturing method of a micro electro mechanical systems (MEMS) switch to which micromachine technology is applied, and a MEMS switch manufactured by the manufacturing method, and more particularly to a MEMS switch having a cantilever structure and used for high-frequency switching.

  2. Description of the Related Art Conventionally, a high frequency switch having a function of switching the passage and reflection of a high frequency signal is used for high frequency electronic devices. In recent years, circuits and electronic devices using MEMS switches to which micromachine technology is applied have been published in academic societies and papers. Note that MEMS switches used for high frequencies are generally classified into shunt switches that repeat the ON / OFF operation when the movable portion moves in parallel with the substrate, and cantilever switches that are cantilevered.

  MEMS switches are implemented by advances in existing semiconductor technology and micromachine technology. The manufacturing equipment can use existing semiconductor equipment, and a fine structure can be formed on the substrate by making use of a fine processing technique or the like. For example, a cantilever type MEMS switch is provided between an input side signal line and an output side signal line. Two opposing electrodes are formed between the input and output signal lines. One electrode is a suction electrode formed at the same surface position as the input and output signal lines, and the other is formed on the suction electrode in a state where a certain interval is maintained (hereinafter referred to as a gap). It is a movable electrode. The movable electrode is provided on a support, such as an anchor, and is held in a suspended state on the suction electrode through a gap of several μm in height. Is required.

The cantilever type MEMS switch is turned ON / OFF by the electrostatic force of the suction electrode. The OFF (reflection) state is a state where no voltage is supplied to the suction electrode (0 v), and the movable electrode and the output side signal line are separated from each other. The ON (passing) state will be described below.
First, by applying a voltage to the suction electrode, a Coulomb force acts on the movable electrode. Next, the movable electrode is elastically deformed and bent to the output side signal line side. The protrusion of the movable electrode comes into contact with the output side signal line.

The steps necessary to obtain such a MEMS switch will be described below.
First, input / output side signal lines and suction electrodes are formed in a lump. Here, they are formed in a necessary region by making full use of a film forming technique such as sputtering, a photoengraving technique and an etching technique. Next, a sacrificial layer is provided in the periphery of the suction electrode and the input and output side signal lines. This sacrificial layer functions as the several micron gap required on the previously described suction electrode. Next, patterning of the sacrificial layer and processing technique are sequentially performed, and after a necessary region is covered with the sacrificial layer, a film serving as a movable electrode is provided. After this is finely processed by patterning and dry etching using a plate making technique, a structure as a movable electrode is obtained. Finally, by removing the sacrificial layer, a cantilever type MEMS switch having a suction electrode and a movable electrode between input and output side signal lines is completed.

  A metal material is generally used for the movable electrode of the MEMS switch. This metal material is generally formed by plating or sputtering. Here, after a seed layer made of a metal material is formed in advance by sputtering or vacuum evaporation, the film is grown by electrolytic plating.

This movable electrode is provided on a sacrificial layer made of a metal material, and the sacrificial layer is not flat because it is covered along the step of the suction electrode formed on the substrate. A movable electrode is formed along the shape.
This is a problem caused by using a sputter or a vacuum deposited film for forming the sacrificial layer, and is caused by poor adhesion of the sacrificial layer to the side wall of the step of the suction electrode. Depending on the formation state, voids may be generated in the film.

Further, the sacrificial layer is processed to form a protrusion of the movable electrode. The protrusion on the output signal line can be obtained by previously drilling a hole in the sacrificial layer and growing the sheet layer and the plating film.
Finally, metal wiring is generally used for the input and output side signal lines. The wiring shown here is generally obtained by processing a film formation such as sputtering and a plate making technique and pattern processing such as dry or wet in order of process.

  In such a MEMS switch, the protrusion of the movable electrode and the output side signal line repeatedly contact and repel (separate) by switching the switch ON / OFF. When the protrusion comes into contact with the output-side signal line, the suction electrode and the movable electrode may come into contact with each other. This is caused by the poor deposition of the sacrificial layer on the side wall of the step in the suction electrode.

  Patent Document 1 describes a MEMS switch using a metal sacrificial layer. In this publication, the suction electrode and the input / output signal line are covered with an insulating film such as a TEOS oxide film and then smoothed by a polishing process to form a metal sacrificial layer and provide an anchor and a movable electrode. That is, Patent Document 1 is characterized in that it includes a step of selectively removing excess protrusions after covering the gap between the suction electrode and the input / output signal line with an insulating film. Further, the reason for removing the step is that the step between the suction electrode and the input / output signal line causes the movable electrode to bend, and it is described that a switch with poor reliability is formed. .

  In Non-Patent Document 1, a similar cantilever type MEMS switch is created using a resist sacrificial layer.

JP 2004-1186 A Ronald A Coutu Jr et al "Selecting metal alloy electric contact materials for MEMS switches", J. Micromech. Microeng. 14 (2004) 1157-1164

The conventional MEMS switch has the following three problems.
First, there is a problem of sticking (sticking) that occurs during the manufacturing process of the MEMS switch. This sticking mainly appears when the MEMS switch is left in the atmosphere for a long period of time, during repeated movement of the switch, and during the manufacturing process of the MEMS switch. The manufacturing process described here refers to a process during or after the sacrifice layer removal.

  In order to prevent the above-mentioned sticking, the conventional technique is improved by using a process such as a supercritical drying process or a freeze drying process, but it is not sufficient. The reason for this is that washing and drying are indispensable for the sacrificial layer removal process, but this process damages the switch. In addition, a switch that has already been fixed in the course of this process is unlikely to return to its original state in subsequent processes. Therefore, no matter how excellent the supercritical drying or freeze-drying process is, there are switches that are already fixed before this process is performed. In addition, supercritical drying and freeze-drying processing take almost half a day, which hinders improvement of processing capacity.

Further, the sticking caused by the repetitive motion of the MEMS switch is caused by temperature change, humidity, and atmosphere due to friction. As a result, an organic contamination source or the like adheres to the contact surface of the movable electrode, and does not leave when contacting the output-side signal line.
As a countermeasure against this sticking, a method of providing a plurality of protrusions on the movable electrode or covering with a hydrophobic film having a low contact angle is employed. However, even when a hydrophobic film or a protrusion is provided, the sticking problem cannot be completely solved, and a process improvement for repairing this problem is required.

The second problem is that an electrode short-circuit occurs between the movable electrode and the suction electrode due to repetitive movement of the MEMS switch, and a part of the suction electrode is dissolved. It is clear that this electrode dissolution is likely to occur in a sacrificial layer using an inorganic material or a metal material, and this is because the wiring pattern step such as the suction electrode is directly reflected in the shape of the sacrificial layer. For example, at the end of the suction electrode, especially at the end of the suction electrode, and at the end of the output signal line, the film does not fit well, and the film tends to be thinner than other flat regions. Moreover, the film | membrane of a movable electrode will wrap around in the clearance gap between a suction electrode and a movable electrode. Thus, since the sacrificial layer has a structure reflecting the suction electrode and the input / output side signal line structure, the movable electrode has a convex portion protruding to the substrate side, and the convex portion comes into contact with the suction electrode. . Due to the contact during energization, the suction electrode dissolves and the movable electrode does not operate.
In addition, there is a problem that the film wraps around due to voids that are easily formed at the end of the electrode, and this film comes into contact with the suction electrode. Furthermore, this phenomenon occurs more significantly when the thickness of the sacrificial layer using a metal material is further reduced.

  Further, as disclosed in Japanese Patent Application Laid-Open No. 2004-1186, a method of flattening by embedding a step between the suction electrode and the input / output side signal line is also conceivable. However, the process is complicated, there is a risk that the signal line and the suction electrode may be simultaneously overetched in addition to the insulating film that is selectively removed, and further, an abrasive is sandwiched between the insulating film and the electrode. Problems such as loss of adhesion of the sacrificial layer are considered, and it is necessary to solve these problems.

  A third problem is an increase in contact resistance due to repetitive movement of the MEMS switch. This change in contact resistance is considered to be caused by contamination of the protrusions of the movable electrode, and becomes more apparent when an organic substance such as a photoresist is used for the sacrificial layer than when an inorganic or metal material is used. The reason for this is that when a movable electrode is formed on an organic sacrificial layer, an altered layer is formed in the resist by gas ions attacking the photoresist surface layer by plasma discharge such as vapor deposition or sputtering. It is considered that the contact resistance is increased by the residue remaining on the protrusion surface. This is an obstacle when using an organic sacrificial layer.

As described above, the conventional MEMS switch has a manufacturing process problem that degrades the operation quality of the switch.
The present invention has been made to solve such problems, and an object of the present invention is to provide a MEMS switch manufacturing method and a MEMS switch that improve switching characteristics and reliability in the MEMS switch as compared with the conventional one. To do.

In order to achieve the above object, the present invention is configured as follows.
That is, the MEMS switch manufacturing method according to the first aspect of the present invention is supported by the first electrode, the first electrode, the suction electrode, and the second electrode that are formed on the substrate and arranged through the gap, respectively. This is a method of manufacturing a MEMS switch including a cantilever-like, flexible, driven by the suction electrode and the second electrode and a movable electrode for switching. First, after forming the first electrode, the suction electrode, and the second electrode on the substrate, a first sacrificial layer is formed on at least a portion of the second electrode. After the formation of the first sacrificial layer, a second sacrificial layer is formed on the suction electrode and at least a portion of the first sacrificial layer. After the formation of the second sacrificial layer, the movable electrode is formed to cover the first sacrificial layer and the second sacrificial layer. Then, the MEMS switch is formed by sequentially removing the first sacrificial layer and the second sacrificial layer.

According to the manufacturing method of the MEMS switch in the first aspect of the present invention, the sacrificial layer provided on the substrate for forming the movable electrode is divided into the first sacrificial layer and the second sacrificial layer made of different materials. did. That is, the first sacrificial layer was formed on at least a part of the second electrode, and after the formation of the first sacrificial layer, the second sacrificial layer was formed on at least a part of the suction electrode and the first sacrificial layer. Further, after the movable electrode is formed, the first sacrificial layer is removed, and finally the second sacrificial layer is removed to manufacture the MEMS switch. With this configuration, the freeze-drying or supercritical drying process that has been conventionally performed for removing the sacrificial layer is not required, and the processing capability can be improved. Even if a cleaning process such as water washing and drying is performed for the removal of the first sacrificial layer, the second sacrificial layer remains at the lower part of the movable electrode at this point, and thus the cleaning process damages the movable electrode. In addition, it is possible to improve the fixing of the movable electrode. Furthermore, since the first sacrificial layer and the second sacrificial layer are different materials, the above-described problems of electrode dissolution and contact resistance increase can be solved by using an appropriate material according to the formation location thereof. it can.
As described above, according to the MEMS switch manufacturing method, the switching characteristics and reliability of the MEMS switch can be improved as compared with the related art.

  A MEMS (Micro Electro Mechanical Systems) switch which is an embodiment of the present invention and a method for manufacturing the MEMS switch will be described below with reference to the drawings. In each figure, the same or similar components are denoted by the same reference numerals.

Embodiment 1 FIG.
1A is a plan view showing a state in which a MEMS switch is loaded on a high-frequency signal line according to Embodiment 1 of the present invention, and FIG. 1B is a perspective view thereof. FIG. 2 is a structural cross-sectional view of the MEMS switch in the AA ′ portion shown in FIG. 1B.
As shown in FIG. 1A, FIG. 1B, and FIG. 2, the MEMS switch 101 according to the first embodiment is an example that functions as a substrate 1, an input-side signal line 2 that functions as a first electrode, and a second electrode. Output side signal line 3, suction electrode 4, and movable electrode 5.

  In the first embodiment, high resistance silicon with a thermal oxide film 13 is used as the substrate 1. The high resistance silicon is used for the purpose of suppressing the generation of parasitic capacitance of the substrate in the high frequency signal or improving the loss to the substrate side. The input side signal line 2 and the output side signal line 3 are high frequency signal lines formed of a metal material on the thermal oxide film 13. Further, for example, a suction electrode 4 made of a metal material is provided on a wiring width of 100 μm. Therefore, the input side signal line 2, the output side signal line 3, and the suction electrode 4 are arranged with appropriate gaps 21 and 22.

  The movable electrode 5 is a cantilever-like flexible electrode whose one end 5a is supported by the input-side signal line 2. The gap 5 extends above the suction electrode 4 and the other end 5b is the output-side signal. It extends to above the line 3. As shown in FIGS. 1A and 1B, at a position near the gap 21, the movable electrode 5 is formed with a hole 17 penetrating the movable electrode 5 along the thickness direction 1 a of the substrate 1, thereby the movable electrode 5. Has a hinge portion 7. Therefore, the movable electrode 5 can be easily bent by the hinge portion 7.

  Such a movable electrode 5 is attracted by an electrostatic force by applying a voltage to the suction electrode 4 and elastically deforms, and the projection 6 provided on the other end 5b of the movable electrode 5 and the output-side signal line 3 come into contact with each other. Turn on (pass). On the other hand, when the voltage application to the suction electrode 4 is stopped, the movable electrode 5 is restored, and the projection 6 and the output side signal line 3 are not in contact with each other and are turned off. In the present embodiment, the protrusions 6 are formed in two places as shown in FIG. 1B, but the number and positions of the protrusions 6 are not specified.

  Next, a manufacturing method of the MEMS switch 101 configured as described above will be described with reference to FIGS. 2 and 3A to 3G.

  First, as shown in FIG. 3A, the input-side signal line 2, the suction electrode 4, and the output-side signal line 3 are formed of a metal material on the substrate 1 on which the thermal oxide film 13 is formed. The type of the metal material is selected on the basis of superiority against alteration such as thermal diffusion and low resistance. In Embodiment 1, for example, a gold sputtered film having a thickness of 500 nm is used. Also, the sputtered film is processed using existing semiconductor equipment.

  That is, first, a resist is applied, and a necessary area is masked through a plate making process including exposure and development with a stepper or the like. Thereafter, pattern processing is performed by wet etching or physical etching. Next, the resist mask is removed by ashing using a resist stripping solution or an ashing apparatus, and necessary electrode shape patterns are formed on the substrate 1.

  Next, as shown in FIG. 3B, a first sacrificial layer 8 is formed in a region on the output-side signal line 3 and located on the suction electrode 4 side. The formation region of the first sacrificial layer 8 on the output side signal line 3 is a region including a portion where the protrusion 6 of the movable electrode 5 is in contact with the output side signal line 3. In the present embodiment, the first sacrificial layer 8 is made of a metal material. The reason is to prevent the protruding portion 6 of the movable electrode 5 and the output-side signal line 3 in contact with the protruding portion 6 from being contaminated with an organic material. In the first embodiment, for example, after using a copper sputtered film having a thickness of about 2.0 μm, the first sacrifice of the structure shown in the region required on the output-side signal line 3 is performed by photolithography and pattern processing. Layer 8 was formed.

Thus, by providing the first sacrificial layer 8 made of a metal-based material corresponding to the region serving as the contact point between the movable electrode 5 and the output-side signal line 3, the region serving as the contact point is contaminated with the organic material. Can be prevented, and the above-described problem of contact resistance change can be solved.
In addition, by forming the first sacrificial layer 8 in a region on the output-side signal line 3 and positioned on the suction electrode 4 side, the second sacrificial layer 8 is formed above the suction electrode 4 as described below. The sacrificial layer 9 can be formed alone. As a result, effects such as “flattening” can be obtained as described below.

  Here, the first sacrificial layer 8 includes the height of the protrusion 6 and a space formed between the protrusion 6 and the output-side signal line 3 (hereinafter referred to as an air gap 16 (see FIG. 2)). A film thickness corresponding to the height plus the height is required. Further, the height of the air gap 16 in the thickness direction 1a of the substrate 1 is set to the same height as the gap 15 between the suction electrode 4 and the movable electrode 5, or the restoring force of the movable electrode 5 is improved. The gap 15 is preferably slightly smaller than the gap 15. In the first embodiment, the height of the protrusion 6 is set to 1 μm, for example, with respect to the film thickness of the first sacrificial layer 8 of 2 μm. Therefore, the air gap 16 is formed with the remaining 1 μm.

  As shown in FIG. 3C, the first sacrificial layer 8 is formed with a first recess 27 for forming the protrusion 6. The first recess 27 is formed by sequentially performing a photoresist mask, ion beam etching, and resist removal.

  Next, as shown in FIG. 3D, from the gap 21 between the input-side signal line 2 and the suction electrode 4, the suction electrode 4, the gap 22 between the suction electrode 4 and the output-side signal line 3, and the first The second sacrificial layer 9 is formed in the sacrificial layer 8 in a range up to just before the first recess 27. In the present embodiment, an organic material is used as the second sacrificial layer 9. This flattens the surface 9a of the second sacrificial layer 9 on the suction electrode 4 regardless of the level difference caused by the input side signal line 2, the suction electrode 4, the output side signal line 3, and the gaps 21 and 22. Because. By the planarization, it is possible to prevent electrode dissolution caused by a short circuit between the movable electrode and the suction electrode accompanying the operation of the switch, which occurs in the conventional MEMS switch.

  Here, the second sacrificial layer 9 is processed on the order of several microns in order to form the gap 15 between the suction electrode 4 and the movable electrode 5 with good control. To reverse the above description regarding the air gap 16, the thickness of the second sacrificial layer 9 must be equal to or greater than the height of the air gap 16 of the first sacrificial layer 8. This is to prevent the central portion 5c of the movable electrode 5 from coming into contact with the suction electrode 4 side. By making the film thickness of the second sacrificial layer 9 larger than the height of the air gap 16, the protrusion 6 is brought into contact with the output-side signal line 3 before the central portion 5 c of the movable electrode 5 comes into contact with the suction electrode 4. be able to. For this reason, in the first embodiment, for example, a photoresist having a thickness of 1.5 μm is used as the second sacrificial layer 9.

  Further, as shown in FIG. 3D, the second sacrificial layer 9 has a region 9b that runs on the first sacrificial layer 8. This is in consideration of the processing accuracy and dimensional tolerance of the second sacrificial layer 9, and is necessary for the reason that no gap is provided between the first sacrificial layer 8. Note that the height of the surface 9c of the region 9b is naturally higher than that of the surface 9a.

  In the conventional MEMS switch, as shown in FIG. 5, the movable electrode has a stepped portion A that protrudes toward the substrate due to the coverage of the sacrificial layer formed on the substrate. On the other hand, by providing the second sacrificial layer 9 with a region 9b that folds over the first sacrificial layer 8 as in the first embodiment, a portion that protrudes from the movable electrode toward the substrate side is not formed. Thus, it can be said that the region 9b is a part that functions to prevent the movable electrode 5 and the suction electrode 4 from contacting each other and to dissolve both of them.

  Next, as shown in FIG. 3E, the second sacrificial layer 9 is provided with a seed layer 14 made of a gold sputtered film, for example, having a thickness of 100 nm, and patterning is performed as shown in the figure. Further, a resist mask 10 is formed in a region of the output side signal line 3 that is not covered with the first sacrificial layer 8. The resist mask 10 is provided as a stopper for separating the movable electrode 5 that is a cantilever from the output-side signal line 3. Gold plating does not grow under the resist mask 10 on the output side signal line 3.

  Next, as shown in FIG. 3F, a material that becomes the movable electrode 5 is formed on the seed layer 14. In the first embodiment, a film is grown on the seed layer 14 using electrolytic gold plating. This gold plating used as the movable electrode 5 requires a film thickness of 5 μm or more in order to obtain the strength of the anchor portion of the movable electrode 5 and the strength of the hinge 7 provided on the movable electrode 5.

  Next, the cross-sectional structure of the switch as shown in FIG. 3F is obtained by sequentially performing pattern processing such as plate making and ion beam. By completing all the steps so far, all the structures having the function of the MEMS switch 101 are formed on the substrate 1. Subsequent steps are mainly performed by removing the first sacrificial layer 8 and the second sacrificial layer 9.

  First, as shown in FIG. 3G, the first sacrificial layer 8 is removed. In the first embodiment, since the first sacrificial layer 8 uses a copper sputtered film, the removal process of the first sacrificial layer 8 includes wet etching using an etchant containing ferric chloride as a component, and washing with water. A drying process is performed. As described above, when the first sacrificial layer 8 is removed, the second sacrificial layer 9 still exists below the movable electrode 5, so that the removal operation of the first sacrificial layer 8 adversely affects the switch structure. There is no.

  Finally, by removing the second sacrificial layer 9 and the resist mask 10 together with a resist carbonization apparatus, the final process of the MEMS switch that becomes the cantilever can be completed by a dry process. Thus, a cantilever type MEMS switch structure can be obtained on the input and output side signal lines 2 and 3 as shown in FIG.

  As shown in FIG. 2, the movable electrode 5 in the MEMS switch 101 formed in this way includes a third recess 23 formed corresponding to the gap 15 by the second sacrificial layer 9, and an output side signal from the gap 22. It has the 4th recessed part 24 which has the depth exceeding the depth of the 3rd recessed part 23 in the thickness direction 1c formed facing the range of the line | wire 3. As shown in FIG.

  As described above, according to the MEMS switch 101 in the first embodiment, the sacrificial layer is formed by being divided into the first sacrificial layer 8 and the second sacrificial layer 9. Thereby, the following effects can be acquired.

  By dividing the sacrificial layer into a plurality of sacrificial layers made of different materials, in the region including the step between the suction electrode 4 and the input and output signal lines 2 and 3, the step is directly formed into a sacrificial layer shape. An organic sacrificial layer that can be planarized without reflection, that is, the second sacrificial layer 9 can be used. Therefore, it is possible to flatten the opposing surface of the suction electrode 4 in the movable electrode 5. As a result, it is possible to prevent a short-circuit between the movable electrode 5 and the suction electrode 4 without embedding the gaps 21 and 22 between the suction electrode 4 and the input and output signal lines 2 and 3, thereby preventing electrode dissolution. Can be prevented. In addition, due to the flattening, the movable electrode 5 does not come into contact with the suction electrode 4 even with the same amount of deflection as in the prior art, so that the thickness of the first sacrificial layer 8 can be reduced. The thickness of the second sacrificial layer 9 that forms the gap 15 on the suction electrode 4 can also be reduced. Therefore, since the movable electrode 5 can be attracted and bent at a lower voltage than before, the voltage of the attracting electrode 4 can be reduced, and the voltage of the circuit using the MEMS switch 101 can be reduced.

  Furthermore, the restoring force of the movable electrode 5 is largely due to the cross-sectional structure of the hinge 7 in the movable electrode 5. If the hinge 7 is twisted or bent, the MEMS switching operation is significantly affected. However, the leveling of the movable electrode 5 on the suction electrode 4 is eliminated by the planarization using the second sacrificial layer 9 as described above. Therefore, the robustness of the MEMS switch 101 is improved.

  Further, the protrusion 6 of the movable electrode 5 where problems such as contact contamination and sticking occur is formed by using the inorganic or metal-based material, that is, the first sacrificial layer 8, and forming the protrusion 6. Since it is covered with the first sacrificial layer 8 made of an inorganic or metallic material, there is no risk of exposure to organic contamination. Therefore, it is possible to prevent problems such as contact contamination, adhesion, and contact resistance increase.

  Furthermore, since the first sacrificial layer 8 and the second sacrificial layer 9 made of different materials are separately formed, they can be removed separately. Therefore, even if wet etching is performed to remove the first sacrificial layer 8, the second sacrificial layer 9 is present below the movable electrode 5 at that time. Therefore, the adhesion during the removal process of the first sacrificial layer 8 can be improved, and even if the washing process such as water washing and drying is performed as the removal process of the first sacrificial layer 8, the MEMS switch is damaged. Never give. Similarly, after the first sacrificial layer 8 is removed, the second sacrificial layer 9 can be finally removed with a low damage asher or the like. A process becomes unnecessary. Therefore, the processing capability can be improved, and a highly accurate MEMS switch can be obtained.

  In the first embodiment, gold is used as the metal material of the suction electrode 4, the movable electrode 5, the input side signal line 2, and the output side signal line 3, but copper, aluminum, platinum, Ru (ruthenium), Rh A low-resistance metal material such as (rhodium) or an alloy thereof may be used. However, it is necessary to select a material that is different from the first sacrificial layer 8 and that is resistant to the etchant solution of the first sacrificial layer 8.

  Further, a material other than the copper used in the first embodiment may be selected for the first sacrificial layer 8 including the seed layer 14. For example, a metal material such as nickel or aluminum can be formed by sputtering or vacuum deposition. Similar effects can be obtained by using TEOS silicon oxide film, silicon nitride film, and amorphous silicon, which are inorganic materials. These films are formed by physical vapor deposition (PVD) such as sputtering or chemical vapor deposition (CVD).

  Further, although a photoresist is used for the second sacrificial layer 9, the present invention is not limited to this. For example, polyimide or SOG (Spin on Glass) may be used by spin coating. However, it is desirable to use a dry etching process to remove the second sacrificial layer 9, and any material may be used as long as it is suitable for this.

  Further, it is desirable that the second sacrificial layer 9 is formed alone on the suction electrode 4. Further, the region 9b where the first sacrificial layer 8 and the second sacrificial layer 9 are folded is provided in the region of the output side signal line 3 or the tip of the output side signal line 3 on the suction electrode side as shown in FIG. 3D. To the output-side signal line side tip of the suction electrode 4. This is because the region of the suction electrode 4 under the movable electrode 5 is flattened to obtain a shape in which dissolution between the movable electrode 5 and the suction electrode 4 does not occur.

Embodiment 2. FIG.
Next, with reference to FIGS. 4A to 4D, the MEMS switch 102 and the manufacturing method thereof according to Embodiment 2 of the present invention will be described. The processes from the process of forming the input / output signal lines 2 and 3 and the suction electrode 4 to the patterning process of the first sacrificial layer 8 shown in FIG. 3B are the same as those of the MEMS switch 101 described above. Therefore, the description here is omitted. The first sacrificial layer, the second sacrificial layer, the movable electrode, and the protrusion in the MEMS switch 102 are the first sacrificial layer 8, the second sacrificial layer 9, the movable electrode 5, and the protrusion 6 in the MEMS switch 101 described above. In the second embodiment, the first sacrificial layer 28, the second sacrificial layer 29, the movable electrode 25, and the second sacrificial layer It numbers with the protrusion part 26. FIG.

  FIG. 4A is a cross-sectional view of the MEMS switch 102 showing a state in which the first sacrificial layer 28 and the second sacrificial layer 29 are patterned on the substrate 1. The difference between the MEMS switch 102 and the MEMS switch 101 described above is that the first sacrificial layer 28 is connected to the suction electrode 4 and the output signal line 3 from the end surface portion 4a of the suction electrode 4 positioned on the output signal line 3 side. And the second sacrificial layer 29 covers almost the entire surface of the first sacrificial layer 28. The second sacrificial layer 29 covers the entire surface of the first sacrificial layer 28. . Here, the second sacrificial layer 29 is formed of an organic resin material provided on the first sacrificial layer 28 in order to form the first recess 27 in the first sacrificial layer 28 for forming the protrusion 26 in the movable electrode 25. Patterning material. That is, the second sacrificial layer 29 serves as both the patterning material and the sacrificial layer. Further, the second sacrificial layer 29 is formed with a second recess 30 for forming the protrusion 26 together with the first recess 27.

  The first sacrificial layer 28 is formed in a range from the end surface portion 4 a to a part of the output signal line 3, so that the second sacrificial layer 29 can be formed alone above the suction electrode 4. It will be possible. Thereby, effects such as “flattening” can be obtained as described above.

FIG. 4A shows a state in which the first sacrificial layer 28 is dug, for example, by ion beam etching or the like. As described above, the protruding portion 26 is formed by etching the first sacrificial layer 28 and the second sacrificial layer 29, and the cross-sectional structure of the protruding portion 26 is determined by an inclination angle at the time of etching. Therefore, the inclination angle formed with the thickness direction 1c of the substrate 1 in the wall surfaces 27a, 30a forming the first recess 27 and the second recess 30 can be different between the first recess 27 and the second recess 30.
It should be noted that the thickness of the second sacrificial layer 29 is considered to be smaller than the initial thickness because it is reduced during the ion beam etching process.

  Next, in order to form the seed layer 14 and the movable electrode 25 shown in FIG. 4B while leaving the second sacrificial layer 29 functioning as a patterning material on the substrate 1, the second sacrificial layer 29 and the like are subjected to a gold plating process and a resist. The formation of the mask 10 is sequentially performed. As described above, the protrusion 26 includes the first recess 27 dug in the first sacrificial layer 28 and the second recess 30 formed of the second sacrificial layer 29. .

Next, as shown in FIG. 4C, the first sacrificial layer 28 is removed by wet etching or the like. Further, as shown in FIG. 4D, the second sacrificial layer 29 and the resist mask 10 and the seed layer 14 remaining therebelow are removed. Is etched back, for example, by ion beam etching.
In this way, the MEMS switch 102 according to the second embodiment can be obtained. Note that the MEMS switch 102 also has the third concave portion 23 and the fourth concave portion 24 as in the MEMS switch 101.

  The MEMS switch 102 according to the second embodiment configured as described above has the same effects as the MEMS switch 101 described above, and can obtain the following specific effects.

  That is, as described above, the protrusion 26 in the movable electrode 25 is formed by the first sacrificial layer 28 and the second sacrificial layer 29. Therefore, the cross-sectional structure of the protrusion 26 is determined by the patterning shape of the second recess 30 in the second sacrificial layer 29 and the shape of the first recess 27 in the first sacrificial layer 28. Therefore, it is possible to have two inclination angles corresponding to the wall surface 27 a of the first recess 27 and the wall surface 30 a of the second recess 30. For this reason, the ion beam etching process requires only a small portion with respect to the first recess 27 that forms the tip of the protrusion 26, and therefore the processing time can be shortened.

  Further, by diverting the second sacrificial layer 29 as a mask for forming the protruding portion 26, a separate resist mask is not required for forming the protruding portion, and the process can be simplified.

  Further, in this MEMS switch 102, the movable electrode 25 is provided while the first sacrificial layer 28 is etched using the second sacrificial layer 29 as a mask. Therefore, an organic substance is formed in the hole for the protrusion 26 provided in the first sacrificial layer 28. There is no intervening.

Note that the conditions used in the first embodiment are applied as they are to the materials and film thicknesses constituting the MEMS switch 102.
Further, as shown in FIG. 4A, the first sacrificial layer 28 is provided up to the gap 22 and the end surface portion 4a of the suction electrode 4 near the output side signal line 3, but this is not restrictive. You may provide in the formation range of the 1st sacrificial layer 8 shown to 3B. However, it is desirable that the second sacrificial layer 29 is provided alone on the suction electrode 4.

In the second embodiment, as described above, the protrusion 26 is formed by the first recess 27 and the second recess 30, but the present invention is not limited to this. As another example of forming the protrusion 26, the first sacrificial layer 28 is not processed, that is, the first recess 27 is not formed, and the process proceeds to the next process, and the second recess 30 is formed in the second sacrificial layer 29. It is also possible to form the protrusion 26 only by doing so. In this case, the cross-sectional structure of the protrusion 26 depends on the pattern shape of the second sacrificial layer 29. Further, the tip of the protrusion 26 is formed on the upper surface 28 a of the first sacrificial layer 28.
By adopting such a configuration, it is not necessary to process the first sacrificial layer 28, so that the process can be simplified.

It is a top view of the MEMS switch in Embodiment 1 of this invention. FIG. 2 is a perspective view of the MEMS switch shown in FIG. 1. It is sectional drawing of the MEMS switch in the A-A 'part shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch shown in FIG. It is a figure explaining the manufacturing process of the MEMS switch in Embodiment 2 of this invention. It is a figure explaining the manufacturing process of the MEMS switch in Embodiment 2 of this invention. It is a figure explaining the manufacturing process of the MEMS switch in Embodiment 2 of this invention. It is a figure explaining the manufacturing process of the MEMS switch in Embodiment 2 of this invention. It is sectional drawing which shows the structure of the conventional MEMS switch.

Explanation of symbols

1 substrate, 1c thickness direction, 2 input side signal line, 3 output side signal line,
4 suction electrode, 5 movable electrode, 6 protrusion, 8 first sacrificial layer,
9 second sacrificial layer, 15 gap, 23 third recess, 24 fourth recess,
27 first recess, 25 movable electrode, 26 projection, 28 first sacrificial layer,
29 second sacrificial layer, 30 second recess,
101, 102 MEMS switch.

Claims (8)

A first electrode, a suction electrode, and a second electrode that are formed on the substrate and arranged via a gap, and are cantilever-like flexible supported by the first electrode and driven by the suction electrode. In a method of manufacturing a MEMS switch including the second electrode and a movable electrode that performs switching,
Forming the first electrode, the suction electrode, and the second electrode on the substrate, and then forming a first sacrificial layer on at least a portion of the second electrode;
After the formation of the first sacrificial layer, a second sacrificial layer is formed on the suction electrode and at least a portion of the first sacrificial layer;
After the formation of the second sacrificial layer, the movable electrode is formed so as to cover the first sacrificial layer and the second sacrificial layer, and then the first sacrificial layer and the second sacrificial layer are sequentially removed to form the MEMS switch. Forming,
A method for manufacturing a MEMS switch, comprising:   The first sacrificial layer is made of an inorganic or metal material, and the first sacrificial layer is provided with a first recess for forming a protrusion provided on the movable electrode and in contact with the second electrode. The method of manufacturing a MEMS switch according to claim 1.   The MEMS switch manufacturing method according to claim 2, wherein the first sacrificial layer is formed in a range from an end surface portion of the suction electrode near the second electrode to the second electrode.   The method of manufacturing a MEMS switch according to claim 2, wherein the second sacrificial layer is formed from above the suction electrode to immediately before the first recess.   The first sacrificial layer is made of an inorganic or metal material, and the second sacrificial layer is a patterning material that is provided on the movable electrode and forms a protrusion that contacts the second electrode. 2. The method of manufacturing a MEMS switch according to claim 1, wherein the second sacrificial layer is formed to cover the entire surface of the electrode and the first sacrificial layer, and the second sacrificial layer is formed with a second recess for forming the protrusion. .   The method for manufacturing a MEMS switch according to claim 5, wherein the first sacrificial layer is formed with a first recess corresponding to the second recess to form the protrusion.   A MEMS switch manufactured by the method for manufacturing a MEMS switch according to claim 1. A first electrode, a suction electrode, and a second electrode that are formed on the substrate and arranged via a gap,
One end supported by the first electrode, extending from above the suction electrode to above the second electrode, a cantilever-like flexible electrode that is driven by the suction electrode and performs switching with the second electrode; With
The movable electrode is formed so as to be opposed to the third recess formed to face the suction electrode and the range from the end surface portion of the suction electrode near the second electrode to the second electrode, and the thickness of the substrate. A MEMS switch comprising a fourth recess formed in a direction with a depth exceeding the third recess.

Images